Prostate cancer (PCa) is a major public health concern for which the substantial morbidity and mortality burden on society is expected to grow with the aging baby boom generation. The carcinogenic effects of estrogen on the prostate have been clearly demonstrated in animal, cell line, and tissue expression studies. Paradoxically, there is a current resurgence of interest in using synthetic estrogens to treat patients with advanced PCa. The dual effect of estrogen appears to be receptor mediated and dependant on steroid pathway interactions. This study will investigate PCa risk and germline genetic variation in genes encoding the estrogen receptors subtypes ESR1 (ESR1) and ESR2 (ESR2), the gene encoding aromatase (CYP19A1) that converts testosterone to estrogen, and the genes that encode estrogen catabolism enzymes (CYP1A1 and CYP1B1). We propose the following primary aims: 1) To comprehensively genotype ESR1, ESR2, CYP19A1, CYP1A1 and CYP1B1 using tag single nucleotide polymorphisms (SNPs) in a population-based case-control study of PCa and estimate relative risks associated with individual SNPs, haplotypes, and within multigenic pathways. 2) To evaluate genotypes association with PCa risk according to measures of disease aggressiveness. We also propose the following secondary aims: (1) To examine obesity as a possible effect modifier in genotype associations with overall PCa risk;(2) To determine whether genotypes are associated with adverse patient outcomes by calculating risk of PCa recurrence/progression and mortality;(3) To measure ESR1 and ESR2 gene expression in solid tumor tissue samples taken from cases undergoing radical prostatectomy and correlate these findings with ESR1 and ESR2 genotypes;and, (4) To investigate associations of genotypes with patient outcomes among men treated with androgen deprivation therapy. To accomplish these aims, we will to build upon an existing population-based study of 1,457 histologically confirmed PCa cases and 1,351 age-frequency matched controls without a history of PCa. The SNPs will be selected using publicly available data to comprehensively cover common haplotype variation within each gene. Prostate cancer risk will be estimated using adjusted unconditional logistic and polytomous regression. Risk of disease recurrence/progression and prostate-specific cancer mortality will be estimated using Cox proportional hazards (PH) regression. Results from this study may provide novel information on how estrogen pathway genes alter risk of PCa and, more importantly, who may be at risk for more aggressive forms of the disease. In addition, this study may identify patient characteristics that interplay with genotypes to increase risk. Lastly, results may serve as pilot data to further explore novel estrogen or anti-estrogen therapeutic approaches.